Also, doesn't an argument on semantics cover a good deal more than a misuse of a particular word? For example, using the same symbol (word) a few times in an argument, to cover a few different meanings in an argument and treating them all as equally true. For example, Chesterton's objection to the claim that we live in a 'vast, cold, dark universe' is based on Semantics, in that he claims this sort of talk freely moves between three meanings: 1) The Universe is much vaster, colder and darker then the patch of earth we inhabit 2) The Universe is a vast, cold and dark universe. That is, more vast, cold and dark then other universes, and 3) The Universe is vast, cold and dark to an extent which is unpleasant. The first is certainly true, but the people he paraphrase move freely into meanings 2 and 3 when discussing our 'cold universe' based on the evidence for meaning 1. Or another example of an argument from Semantics is to argue that the argument has a symbol, but no substance. I remember I once framed, with Plotinus, that omnipotence is confined by logic in these terms. You may ask "Could an Omnipotent being create a square circle?" and this sounds like a perfectly reasonable question by the rules of English grammar. But in your sentence, "Square circle" doesn't signify anything. You can't tell me the shape you are referring to as a square circle, and so the question is meaningless. It's not even a question. So argument from Semantics is certainly more then a just the correction someone using the wrong definition of a word, it can frequently be a very complete rebuttal of an argument, and not just one on technical grounds.